1890.] 3 



mammilla-like development in the subdorsal and supraventral regions, 

 each mammilla being tipped with a papilla, which can be protruded or 

 retracted at the will of the animal. The object o£ this remarkable 

 configuration is, I believe, to give a secure foothold whilst the animal 

 is feeding at the edge of its hollow or balloon-like mine ; and in 

 support of this explanation is the fact, that a very similar formation 

 of these segments is present in at least one or two other Lepidopterous 

 larvae, in which the conditions are also similar. The abdominal seg- 

 ments are cylindrical, and terminate behind in a fine styliform extremity. 

 Segment five is provided with a pair of lateral protuberances, of little 

 or no use apparently, except perhaps to serve as the distinctive mark 

 of a Micropteryx larva, for the benefit of the naturalist. 



Such are the characters of the mature larva, and with one excep- 

 tion they are equally true of the immature one. In the former, as 

 has already been shown, the three thoracic segments are of equal 

 width, but in the latter or immature larvae this equality does not 

 obtain, segment two being now distinctly wider than either of the 

 others, and consequently the widest segment in the body. This pre- 

 ponderance of segment two is, I believe, characteristic of all mining 

 larvae that form blotches, so long at least as the blotch remains flat, 

 and does not become hollow or balloon-like ; but as soon as the blotch- 

 life is over, as happens half-way in the life of the Gracilariidce, or the 

 blotch becomes hollow and capacious, as in the later life of the Litho- 

 colletes, then segment two loses its superiority, and may even become 

 narrower than the following segment. This being the principle under- 

 lying the development of segment two, we find, as might be expected, 

 that the superiority of segment two is lost early, namely, at the time 

 of the second moult, in those Micropteryges in which the mine from 

 starting near the edge quickly becomes balloon-like, whilst it is re- 

 tained up to and even beyond the last moult in Sparmannella and 

 salopiella, where, from its position in the middle of the leaf, the mine 

 remains flat and shallow until the last few days of rapid excavation. 



I now come to the individual larvae, and will take together three 

 species with exactly similar habits, that begin to feed very early in the 

 year, at the end of April or the beginning of May, before the birch 

 leaves are fully grown. They are unimaculella, semipurpurella, and 

 what is probably a new species, which I will call provisionally incon- 

 spicuella, from the inconspicuousness of the anal wing-spot in the 

 perfect insect. All start from the edge of the leaf with a short 

 twisting gallery, that is soon lost in the after-formed blotch. 



